* Lukáš Czerner <lczerner@xxxxxxxxxx>:
> On Mon, 23 Mar 2015, Eric Whitney wrote:
>
> > Date: Mon, 23 Mar 2015 19:15:39 -0400
> > From: Eric Whitney <enwlinux@xxxxxxxxx>
> > To: Lukáš Czerner <lczerner@xxxxxxxxxx>
> > Cc: Eric Whitney <enwlinux@xxxxxxxxx>, linux-ext4@xxxxxxxxxxxxxxx,
> > tytso@xxxxxxx
> > Subject: Re: [PATCH] ext4: don't consume reserved space when allocating
> > partial cluster
> >
> > * Lukáš Czerner <lczerner@xxxxxxxxxx>:
> > > On Mon, 16 Mar 2015, Eric Whitney wrote:
> > >
> > > > Date: Mon, 16 Mar 2015 21:20:09 -0400
> > > > From: Eric Whitney <enwlinux@xxxxxxxxx>
> > > > To: linux-ext4@xxxxxxxxxxxxxxx
> > > > Cc: tytso@xxxxxxx
> > > > Subject: [PATCH] ext4: don't consume reserved space when allocating partial
> > > > cluster
> > > >
> > > > When xfstests' auto group is run on a bigalloc filesystem with a
> > > > 4.0-rc3 kernel, e2fsck failures and kernel warnings occur for some
> > > > tests. e2fsck reports incorrect iblocks values, and the warnings
> > > > indicate that the space reserved by delayed allocation is being
> > > > overdrawn at allocation time.
> > > >
> > > > Some of these errors occur because the reserved space is incorrectly
> > > > decreased by one cluster when ext4_ext_map_blocks satisfies an
> > > > allocation request by using an unused portion of a previously allocated
> > > > cluster. Because a cluster's worth of reserved space was already
> > > > removed when it was first allocated, it should not be removed again.
> > >
> > > Hi Eric,
> > >
> > > I am not sure I understand. What do you mean by saying that the
> > > space was already removed when it was first allocated ?
> >
> > Hi Lukas:
> >
> > I'm sorry that was confusing - I didn't get the terminology quite right,
> > given the usage in the code. What I'm discussing in that sentence is
> > the space reserved for delayed allocation. Instead of "removed", I should
> > have said "released". If we're mapping from an existing cluster, at some
> > point in the past that cluster was allocated, and at that time the space
> > reservation for that cluster would have been released. So, we ought not
> > to be releasing its space again.
> >
> > >
> > > From my point of view the ext4_da_update_reserve_space() call is ok,
> > > because we're going to use blocks from already allocated cluster, so
> > > we do not want to account for quota in this case, because that has
> > > already been done when the cluster was allocated. The rest is just
> > > updating reservations and the dirty clusters counter which needs to
> > > be done in any case. But I might be actually missing something, am I
> > > ?
> >
> > I agree that we don't want to account for quota, as that should have been
> > done in the past when the cluster was first allocated. I think we don't
> > want to update the reservations or the dirty clusters counter because that
> > should also have been taken into account at the same time in the past. If
> > we update them again, decreasing them once more for the cluster we're currently
> > processing, we'll be double accounting for the space.
> >
> > The code in ext4_da_map_blocks() that runs at write begin time and increases
> > the amount of reserved space only does so when a cluster has not been
> > previously allocated or already accounted for as part of a delalloc extent
> > recorded in the status tree. I think it should be accurately reflecting the
> > number of clusters we'll eventually need to allocate for data, so there's no
> > room for double counting when mapping from an existing cluster in
> > ext4_ext_map_blocks().
>
> Ah, you're right so that's there probably from the times where we
> had to do metadata blocks reservation as well. Thanks for
> clarification.
>
> >
> > If I'm not reading the delalloc accounting code incorrectly, a few more patches
> > will likely be required to remove some of the code immediately following
> > if (!map_from_cluster) and a chunk in ext4_ext_handle_unwritten_extents().
>
> Maybe it would be worth deal with that mess in the
> ext4_ext_map_blocks() within a single patch ?
>
Ideally, yes, but we'd like to get out the safest fix (this one) quickly
so the developers working on SMR support get cleaner bigalloc test runs.
The others will follow. A patch fixing delalloc accounting in
ext4_ext_handle_unwritten_extents() just needs more testing, but
potentially affects non-bigalloc file systems, so is slightly higher risk.
A second patch would remove the code in ext4_ext_map_blocks following the
"if (!map_from_cluster)" statement inserted by this patch. However, it's
actually protecting us from a bug caused by the use of
filemap_write_and_wait_range calls at the top of ext4_punch_hole, etc. -
it causes the same reserved space to be released twice. There's a two year
old patch, never landed, that might be useful to fix this by removing the
filemap call, but it's going to take more time to get that sorted out.
I'll post a v2 that cleans up my terminology problems with a slightly
different title shortly.
Thanks for looking this over!
Eric
> -Lukas
>
> >
> > Thanks,
> > Eric
> >
> >
> > >
> > > Thanks!
> > > -Lukas
> > >
> > > >
> > > > This patch appears to correct the e2fsck failure reported for
> > > > generic/232 and the kernel warnings produced by ext4/001, generic/009,
> > > > and generic/033. Failures and warnings for some other tests remain to
> > > > be addressed.
> > > >
> > > > Signed-off-by: Eric Whitney <enwlinux@xxxxxxxxx>
> > > > ---
> > > > fs/ext4/extents.c | 14 +-------------
> > > > 1 file changed, 1 insertion(+), 13 deletions(-)
> > > >
> > > > diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
> > > > index bed4308..554190e 100644
> > > > --- a/fs/ext4/extents.c
> > > > +++ b/fs/ext4/extents.c
> > > > @@ -4535,19 +4535,7 @@ got_allocated_blocks:
> > > > */
> > > > reserved_clusters = get_reserved_cluster_alloc(inode,
> > > > map->m_lblk, allocated);
> > > > - if (map_from_cluster) {
> > > > - if (reserved_clusters) {
> > > > - /*
> > > > - * We have clusters reserved for this range.
> > > > - * But since we are not doing actual allocation
> > > > - * and are simply using blocks from previously
> > > > - * allocated cluster, we should release the
> > > > - * reservation and not claim quota.
> > > > - */
> > > > - ext4_da_update_reserve_space(inode,
> > > > - reserved_clusters, 0);
> > > > - }
> > > > - } else {
> > > > + if (!map_from_cluster) {
> > > > BUG_ON(allocated_clusters < reserved_clusters);
> > > > if (reserved_clusters < allocated_clusters) {
> > > > struct ext4_inode_info *ei = EXT4_I(inode);
> > > >
> >
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
